1. 1/25 Bandwidth Request Channel IEEE 802.16 Presentation Submission Template (Rev. 9) Document Number: C80216m-09/0015 Date Submitted: 2009-01-05 Source: Sungho Park (Park_SH@lge.com)Park_SH@lge.com Jinyoung Chun (jychun03@lge.com)jychun03@lge.com Bin-Chul Ihm (bcihm@lge.com) LG Electronics * http://standards.ieee.org/faqs/affiliationFAQ.html Re: “802.16m SDD text”: IEEE 802.16m-08/052, “Call for Comments on Project 802.16m System Description Document (SDD)” Target topic: 11.9 UL Control Structure Base Contribution: None Purpose: Discussion and adoption for 802.16m SDD Notice: This document does not represent the agreed views of the IEEE 802.16 Working Group or any of its subgroups. It represents only the views of the participants listed in the “Source(s)” field above. It is offered as a basis for discussion. It is not binding on the contributor(s), who reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE 802.16. Patent Policy: The contributor is familiar with the IEEE-SA Patent Policy and Procedures: and.http://standards.ieee.org/guides/bylaws/sect6-7.html#6http://standards.ieee.org/guides/opman/sect6.html#6.3 Further information is located at and.http://standards.ieee.org/board/pat/pat-material.htmlhttp://standards.ieee.org/board/pat

2. 2/25 Contents Bandwidth Request Procedure –Basic 3-step quick procedure –5-step procedure as fallback mode of 3-step procedure Bandwidth Request Channel Structure –Green Field –Legacy Support Simulation Results –Parameters –Link performances Proposed Text Appendix –Simulation Method –Performance Metric for Indicator –Performance Metric for Message –BR Rate

3. 3/25 Fail Bandwidth Request Procedure Preamble only detection & message error Preamble detection & message detection DLULDLULDLULDLULDLULDLUL 12 345 Transmits BW-REQ with preamble and message UL grant after BW-REQ message -Grant type 1 -Rx info: Rx location, code, partial MS_ID, etc. -Resource allocation info. for data transmission 1 2 UL grant after BW-REQ preamble -Grant type 2 -Rx info: Rx location, code -Resource allocation info. for MAC BW-REQ message (ex. Non-contention BR, BR header) Try BW-REQ after back-off time UL data transmission with Full MS_ID or additional BW-REQ 3 MAC BW-REQ message - Full MS_ID, Flow_ID, Buffer size, etc UL grant after MAC BW-REQ message -Grant type 3 -Full MS_ID, Buffer size, etc -Resource allocation info. for data transmission 2 3 4 5 UL data transmission w/ or w/o additional BW- REQ

4. 4/25 Bandwidth Request Channel Structure Green Field –Bandwidth Request Channel A BW-REQ channel consists of 3 distributed BW- REQ tiles. A BW-REQ tile is defined as 6 contiguous subcarriers by 6 OFDM symbols. –Preamble It can include indicator to notice the ABS of a UL grant request. Allocation –CDM based multiplexing with orthogonal sequence of length 12 –Dimension : 2x6 (basic option) Contents –Partial MS-ID masked with sequence (3 bits) –Data It can include information about the status of queued traffic at the AMS such as buffer size and quality of service, including QoS identifiers Allocation –CDM based multiplexing –Same sequence with Indicator Contents (total 6 bits) –Residual (partial) MS-ID (2~4 bits) –QoS identifier (2~3 bits) –Buffer Size (0 ~ 2 bits) BW-REQ message : 6 Symbols : spread by identical sequence (length 12) adopted in the BW-REQ Indicator : Length 12, : same sequence for three tiles S1S1 S2S2 S3S3 S4S4 S5S5 S6S6

5. 5/25 Bandwidth Request Channel Structure Legacy support mode –Bandwidth Request Channel A BW-REQ channel consists of 6 distributed BW- REQ tiles. A BW-REQ tile is defined as 4 contiguous subcarriers by 6 OFDM symbols –Preamble It can include indicator to notice the ABS of a UL grant request Allocation –CDM based multiplexing with orthogonal sequence of length 12 –Dimension : 2x6 (basic option) Contents –Partial MS-ID masked with sequence (3 bits) –Data It can include information about the status of queued traffic at the AMS such as buffer size and quality of service, including QoS identifiers Allocation –CDM based multiplexing –Same sequence with Indicator Contents (total 6 bits) –Residual (partial) MS-ID (2~4 bits) –QoS identifier (2~3 bits) –Buffer Size (0 ~ 2 bits) BW-REQ message : 6 Symbols : spread by identical sequence (length 12) adopted in the BW-REQ Indicator : Length 12, : same sequence for three tiles S1S1 S2S2 S3S3 S4S4 S5S5 S6S6

6. 6/25 Simulation Parameters CommonItemsValues Simulation Environments & Assumption for BRCH Antenna Configuration1Tx / 2Rx Channel ModelPB3, VA60, VA120 # of BRCH1 BRCH structureThree 6x6 tiles # of MSs1 or 2 False-Alarm Definition 1. Some signals are detected when no signal is transmitted 2. Wrong signals are detected when some signals are transmitted LGEItemsValues BW-REQ Indicator Sequence TypeDFT Sequence Length12 * 3 (same sequence for three tiles) Target False Alarm0.1% Sequence AllocationNo-overlapping Sequence Selection / Random Sequence Selection Sequence DetectionNon-coherent BW-REQ message Information Size6 bits Channel CodingBlock Code (6, 12) ModulationQPSK Allocation MethodCDM (identical sequence with indicator) Channel Estimation2-D MMSE Receiver TypeML Msg Decoding MethodDecode Msg for detected preamble

7. 7/25 Simulation Parameters INTELItemsValues BW-REQ Indicator Sequence TypeDFT Sequence Length19 * 3 (same sequence for three tiles) Target False Alarm0.1% Sequence AllocationNo-overlapping Sequence Selection / Random Sequence Selection Sequence DetectionNon-coherent BW-REQ message Information Size12 bits Channel CodingBlock Code (12, 30) + repetition ModulationBPSK Allocation MethodCSM Channel Estimation2-D MMSE Receiver TypeML Msg Decoding MethodDecode Msg for detected preamble

8. 8/25 Link Curve ( False-alarm Definition 1) Ped A (3km/h) –Threshold Based on False-alarm Definition 1 Target FA : 0.1% –Indicator Detection Target MD Probability : 1% The performance of LGE’s scheme outperforms about 1.5dB, because it can utilize even the data region adopted same sequence. –Message Detection For the case of one user, LGE’s scheme achieve at about -2.8dB, while Intel’s scheme achieve 1% BLER at about -1.2dB. (≈ 1.6dB gain) For two user case, LGE’s scheme achieve at about -2.8dB, while Intel’s scheme achieve 1% BLER at about -0.8dB. (≈ 2dB gain) LGE’s scheme has about 1.6 dB gain for the single user case, and about 2dB gain for two user case. In this simulation, LGE’s scheme achieve the relative gain due to short information size while Intel’s scheme doesn’t include false- alarm effect and high muxing order.

9. 9/25 Link Curve ( False-alarm Definition 1) Veh A (60km/h) –Threshold Based on False-alarm Definition 1 Target FA : 0.1% –Indicator Detection Target MD Probability : 1% The performance of LGE’s scheme outperforms about 1.7dB. –Message Detection For the case of one user, LGE’s scheme achieve at about -1.2dB, while Intel’s scheme achieve 1% BLER at about 0.6dB. (≈ 1.8dB gain) For two user case, LGE’s scheme achieve at about -1.2dB, while Intel’s scheme achieve 1% BLER at about 1.0dB. (≈ 2.2dB gain) LGE’s scheme has about 1.8 dB gain for the single user case, and about 2.2dB gain for two user case. In this simulation, LGE’s scheme achieve the relative gain due to short information size while Intel’s scheme doesn’t include false- alarm effect and high muxing order.

10. 10/25 Link Curve ( False-alarm Definition 1) Veh A (120km/h) –Threshold Based on False-alarm Definition 1 Target FA : 0.1% –Indicator Detection Target MD Probability : 1% The performance of LGE’s scheme outperforms about 1.7dB. –Message Detection For the case of one user, LGE’s scheme achieve at about -1.1dB, while Intel’s scheme achieve 1% BLER at about 0.8dB. (≈ 1.9dB gain) For two user case, LGE’s scheme achieve at about -1.0dB, while Intel’s scheme achieve 1% BLER at about 1.0dB. (≈ 2.0dB gain) LGE’s scheme has about 1.9 dB gain for the single user case, and about 2.0dB gain for two user case. In this simulation, LGE’s scheme achieve the relative gain due to short information size while Intel’s scheme doesn’t include false- alarm effect and high muxing order.

11. 11/25 Link Curve ( False-alarm Definition 2) Ped A (3km/h) –Threshold Based on False-alarm Definition 2 Target FA : 0.1% –Indicator Detection Target MD Probability : 1% The performance of LGE’s scheme outperforms about 1.5dB. –Message Detection For the case of one user, LGE’s scheme achieve at about -3.1dB, while Intel’s scheme achieve 1% BLER at about -1.4dB. (≈ 1.7dB gain) For two user case, LGE’s scheme achieve at about -3.0dB, while Intel’s scheme achieve 1% BLER at about -1.0dB. (≈ 2.0dB gain) LGE’s scheme has about 1.7 dB gain for the single user case, and about 2.0dB gain for two user case. In this simulation, LGE’s scheme achieve the relative gain due to short information size while Intel’s scheme doesn’t include false- alarm effect and high muxing order.

12. 12/25 Link Curve ( False-alarm Definition 2) Veh A (60km/h) –Threshold Based on False-alarm Definition 2 Target FA : 0.1% –Indicator Detection Target MD Probability : 1% The performance of LGE’s scheme outperforms about 1.9dB. –Message Detection For the case of one user, LGE’s scheme achieve at about -1.3dB, while Intel’s scheme achieve 1% BLER at about 0.7dB. (≈ 2.0dB gain) For two user case, LGE’s scheme achieve at about -1.3dB, while Intel’s scheme achieve 1% BLER at about 1.2dB. (≈ 2.5dB gain) LGE’s scheme has about 2.0 dB gain for the single user case, and about 2.5dB gain for two user case. In this simulation, LGE’s scheme achieve the relative gain due to short information size while Intel’s scheme doesn’t include false-alarm effect and high muxing order.

13. 13/25 Link Curve ( False-alarm Definition 2) Veh A (120km/h) –Threshold Based on False-alarm Definition 2 Target FA : 0.1% –Indicator Detection Target MD Probability : 1% The performance of LGE’s scheme outperforms about 1.9dB. –Message Detection For the case of one user, LGE’s scheme achieve at about 0.2dB, while Intel’s scheme achieve 1% BLER at about 3.0dB. (≈ 2.8dB gain) For two user case, LGE’s scheme achieve at about 0.5dB, while Intel’s scheme achieve 1% BLER at about 3.5dB. (≈ 3.0dB gain) LGE’s scheme has about 2.8 dB gain for the single user case, and about 3.0dB gain for two user case. In this simulation, LGE’s scheme achieve the relative gain due to short information size while Intel’s scheme doesn’t include false- alarm effect and high muxing order.

14. 14/25 Link Curve Analysis There are two kinds of link performance according to false- alarm definitions. –LG’s scheme outperforms Intel’s.

15. 15/25 Proposed texts

16. 16/25 Proposed texts (Cont’d)

17. 17/25 Appendix Simulation Criteria Performance Metric for Indicator Performance Metric for Message Threshold & False-alarm Probability BR Rate

18. 18/25 Simulation Criteria Link-level Simulation  Mandatory [Link Reliability – Initial Comparison] –# of Users : 1 or 2 –Sequence Allocation Random Selection –Channel Estimation : 2-D MMSE –Message Detection : MLD [Latency] –Cell Configuration : Single Cell –Traffic Model : VoIP –# of Users : 500 per cell (sector) BR according to VoIP model Mean Talk Spurt = 2.5 sec –Sequence Allocation Random Selection –Full Bandwidth Request Operation Retry, Frame Delay, Fallback Mode No Process Delay Channel Estimation : 2-D MMSE Message Detection : MLD System-level Simulation  Optional –Throughput (w. cell coverage) –Cell Configuration : 57 sectors (2 tiers) –Traffic Model : VoIP BR & Real VoIP Traffic –# of Users : 500 per sector BR according to VoIP model Mean Talk Spurt = 2.5 sec –Sequence Allocation Random Selection –Full Bandwidth Request Operation Retry, Frame Delay, Fallback Mode No Process Delay Channel Estimation : 2-D MMSE Message Detection : MLD –Power Control (?) –Scheduling (?) –Message False-alarm

19. 19/25 Performance Metric for Indicator Definition of False Alarm –Receiver detects unwanted BR whether there are some BR signals or no BR signal. –False Alarm  wrong sequence detection  resource waste –False Alarm depends on cross correlation properties of multiplexing sequences as well as channel selectivity & mobility –Target False Alarm=0.1% (ref. Ranging) Threshold –Fixed threshold Fixed one threshold per SNR level regardless Channel Selectivity and mobility (  AWGN) Fixed one threshold for all SNR but different for channel selectivity and/or mobility –Adaptive threshold SNR Channel Selectivity User Mobility Misdetection Probability –Receiver can’t detect BR –Target Misdetection Probability=1% (ref. Ranging)

20. 20/25 Performance Metric for Message Message Error –False Alarm regarding performance metric Link-level –Ideal message false alarm detection System-level only –Threshold : power level, channel quality (e.g. RSSI, CSI) –Metric Value > Threshold »message error  False Alarm –Metric Value < Threshold »message error  turn into Regular Access –BLER All cases of message error including Indicator error –Indicator detection fail »misdetection –Indicator detection fail but message error »indicator collision, message error

21. 21/25 Threshold & False-alarm Probability Ped A (3km/h)

22. 22/25 Threshold & False-alarm Probability Veh A (60km/h)

23. 23/25 Threshold & False-alarm Probability Veh A (120km/h)

24. 24/25 BR Rate Initial BR Rate –Traffic Model : VoIP –# of VoIP user : 500 / 10 MHz (refer SRD 8.3 (Max  Indoor)) –Mean Talk Spurts 50% Voice Activity  1.25 sec. Bi-directional : 2.5 sec. (refer EMD)  Exponential Distribution with mean = 2.5 sec. –Mean # of BW-REQ/Frame/Sector BR Rate = K VoIP users/sector * N BR/frame = KN BR/Frame/sector 0.4BW-REQ per 200 frames  BW-REQ/500frame 500users/sector * BW-REQ/500frame = 1 BW-REQ/Frame/Sector –Mean Call Time : 180 sec. –# of opportunity / Frame = 1 Timer –Packet Access Delay Boundary : 50 ms Back-off –Random Back-off [0, 40ms]

25. 25/25 BR Rate (Cont’d) VoIP Packet Transmission 20ms Mean 1.25 sec ActiveInactive Mean 1.25 sec 50% Voice Activity Bandwidth Request VoIP Packet Call Duration - Call Duration : Lognormal Distribution based on K-L Divergence Method ¶ ¶ J. Guo, F. Liu, and Z. Zhu, “Estimate the call duration distribution parameters in GSM system based on K-L divergence method,” in Proceedings of the International Conference on Wireless Communications, Networking and Mobile Computing (WiCom 2007), 2007, pp. 2988–2991.